Wan AiO.All speedups.Subgraphs. +FunControl 2.2

v3.0+ requires frontend of 1.25.2 or newer for subgraphs to work.

pip install -U comfyui_frontend_package

Subgraphs have some issues described in full wflow, and how to reduce their impact.

Optimisations

To use torch.compile, you need triton.

pip install triton-windows

To install sageattention2.2.0, go to https://github.com/woct0rdho/SageAttention/releases/tag/v2.2.0-windows, locate version for your comfy build - both python and pytorch versions can be located in Settings->About, - then copy link of fitting wheel and use it with command:

pip install wheel_link

obviously replace wheel_link with the real one.

If it is too hard for you, run

pip install sageattention

It will install older 1.0.6, which is significantly slower than 2.2.0, but still better than nothing.

You need to have MSVC tools and CUDA 12.8 installed and set in PATH.

For portable also grab include_libs.zip from https://github.com/woct0rdho/triton-windows/releases/tag/v3.0.0-windows.post1 appropriate for your python version and unzip it into python_embeded folder.

To gain free easy speed bost, verify that you have pytorch 2.7.0+, if not - update using command

pip install -U torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu128

then in your run.bat file add '--fast fp16_accumulation' parameter.

Fresh Comfy installs are shipped with appropriate torch version, but you need to add parameter manually, and portable has run_nvidia_gpu_fast_fp16_accumulation.bat.

These three speedups together grant ~60-70% boost. Using lightx2v/lightning loras means further x2 boost from 1cfg and x5 from lower number of steps (4 instead of 20).

First launch

Open full+examples workflow and install missing nodepacks.
It contains exposed inner workflows of subgraphs and examples showing how to use and chain subgraphs.
Open wan aio workflow. Change modelpaths for checkpoints you're planning to use, download and select speed-up loras. Save subgraphs as templates for convenient use.

A collection of workflows turned into subgraphs to be used in a single space - to avoid constant jumping between workflows, saving/uploading videos and swapping values.

A workplace, rather than a workflow.

Generate and extend videos with different lengths and loras indefinitely, create loops, upscale.

Workflow seems massive and complex, until you realize you only have to use few nodes.

Use it, prune it, break it into pieces or just learn it's concepts.

What's in

AiO:

• basic functions - T2V, I2V, FLF2V - all with separate V2V option

• FunControl 2.1 and 2.2 - to restyle video based on input image and it's ControlNet

• Wrapper V2V/IV2V nodes - makes it possible to upscale to resolutions unavailible with native nodes

• VACE2.1 anyframe to video - multiple keyframes at various locations

• VACE2.1 seamless motion - extension, loop, infinite video generation

Wrapper discontinued since 2.8. Too many features are being added for me to keep up. May return to it when 2.2 Phantom and/or VACE released.

Changelog

v3.1 17.08.2025

Added:

• FunControl 2.2 - inner wflow and subgraph

Changed:

• How loras added; initially i opted to using lora stacks, as only 2.1 loras were availible, to select lora once and apply to both ckpts, but now it's no longer relevant, so back to PowerLoraLoader

FunInpaint was very disappointing, so i won't be adding it. I2V is same in function and has no risk of conflicts with loras.

v3.0 01.08.2025

Added:

• Wan2.2 model profiles and new nodes utilizing it's MoE capabilities

Changed:

• Group nodes replaced with subgraphs

Removed:

• 1.3b control loras

v2.8 14.07.2025

Added:

• VACE seamless extension - wflow and group node

• VACE seamless loop - wflow and node

Removed:

• HunYuan

• LTX

v2.7 17.05.2025

WanWrapper workflow updated

• VACE 14b and 1.3b

• Phantom

• FantasyTalking

• CausVid (3step) lora

• Spline dot control example

v2.6 24.04.2025

• Wrapper IV2V node - crank up blocks to swap and upscale to XL resolution

• FLF2V - official start-to-end frame model that rocks

• Skyreels V2 DF - video extention model that takes last frames of provided video and seamlessly extends it - pretty much FramePack, but Wan

v2.5 7.04.2025

Wan wrapper consolidated - separate workflows are likely better, keeping it as proof of concept

v2.1. 4.04.2025

InterpV2V nodes added.

v2. 3.04.2025

HunYuan - standard tools: T2V, V2V, I2V (including LeapFusion), IV2V (performs poorly)

introduction of Group Nodes

v1. 31.03.2025

initial release